Dilemmas of structural adjustment and environmental policies under instability

(1)

Pergamon Copyright 0 1996 Elsevier Science Ltd Printed in Great Britain. All tights reserved

0305-750X/96 $15.00 + 0.00 0305-750x(95)00140-9

Dilemmas of Structural Adjustment and

Environmental Policies Under Instability:

Post-1980 Turkey

KORKUT BORATAV

University of Ankara, Turkey

OKTAR TiiREL

Middle East Technical University, Ankara, Turkey

and

EtiNC YELDAN

Bilkent University, Ankara, Turkey

Summary. - The Turkish structural adjustment since 1980 has been associated with chronic instability. Since the late 1980s. the weaknesses in the fiscal system and the premature external liberalization emerge as the main factors hindering the passage toward stable growth. Enforced and erratic distributional changes and relative stagnation of capital accumulation have undermined the growth potential of the economy. Fur- ther, it is demonstrated that existing market structures may negate environmental policies based on market incentives. These observations, as well as those on the interactions of the market system and the environment, create strong arguments in favor of an active state.

1. RECENT ECONOMIC HISTORY

(a) A synopsis of reforms and theirphasing At the time of the writing of this paper (September 1994), the Turkish economy had not yet recovered from a major financial crisis culminating between December 1993 and May 1994, with severe repercus- sions on the real side of the economy. Although destabilizing forces leading to financial misfortunes beginning in 1989 and extremely buoyant economic activity in 1993 gave signs of an impending breakdown, this paper excludes this recent phase, since it is too early to undertake a comprehensive analysis. The study covers 1981 to 1992-93; the earlier crisis of 1978-79 as well as the following year characterized by an orthodox shock treatment and the consequent relative price realignment are excluded. For purposes of comparison with the precrisis years of the 1970s either 1976 or 1977 is taken as a point of reference.

Economic policies followed during 1981-1992 did not deviate from the basic philosophy of orthodox structural adjustment, but the same cannot be argued with respect to conventional stabilization policies, particularly during the second half of the period. On

the basis of policy shifts, three subperiods, i.e. 1981-83,1984-88 and 1989-92 are delineated (Table 1). There are some peculiar features of 1988 in terms of quantitative indicators since this year saw a final (and unsuccessful) attempt at fiscal and monetary restraint with a view to correcting macro imbalances which had been gaining momentum from 1986 onwards (see Table 2). With this qualification in mind, the determining features of the three subperiods are as follows:

- 1981-83 is the “military phase” characterized by enforced incomes policies, with emphasis on stabilization as a sequel to the shock treatment of 1980, deregulation of internal commodity markets and clumsy and disastrous moves into financial liberalization.

- 1984-8 (with the exception of 1988) can be labeled the “golden years” of Motherland Party (MP) and its then leader, Mr. ozal. Decisive steps were taken in this subperiod toward trade and financial liberalization. The emphasis was put on structural adjustment and the necessary institution- building. Starting in 1985, expansionary fiscal policies and a public investment boom led to a vir- tual neglect of stabilization objectives. On the

(2)

other hand, the institutional and legal instruments inherited from the military phase continued to be used effectively to control labor incomes

- The third phase signifies a “return to populism” plus external financial liberalization.” A wage explosion accompanied by higher agricultural prices led to further worsening of public sector finances. Privatization began to be seen as a substi- tute for fiscal reform. Short-term capital movements in response to fluctuations in exchange and interest rates created erratic changes in external accounts.’

A few supplementary remarks on the indicators given in Table 2 may be useful:

- Once the initial slack in the economy which In the context of a three-gap growth model (see, inter alia, Bacha 1990) the saving constraint turns out to be the least binding in the first subperiod. For the successive subperiods, fiscal and foreign exchange (FX) constraints respectively seem to be the least binding.2

contributed to a relatively rapid pace of recovery

- Neither current nor constant price calculations and growth in the first two subperiods was exhausted, the economy settled down to slower growth. The potential growth rate of the Turkish

of average savings and investment propensities economy moved to a lower plateau during the

indicate a significant and permanent improvement 1980s and the early 1990s. A 4.1% p.a. of potential GDP growth rate for the period under study com- pares poorly with the growth path of earlier decades.

-The aftermath of MP’s boom is characterized by the shortening of the growth cycle. The post- 1987 growth average of 3.7% p.a. is made up of two above-trend performances in 1990 and 1992 and three years of declining per capita income (1988-89 and 1991). 1993 and 1994 also seem to fit well into this pattern of demand-led boom and the bust which sets in shortly after the ceiling of potential output is reached.

Table 1. Subperiods based on economicpolicy shifs, 1981-92

Policies 1981-83 1984-88 1989-92 Financial system Fiscal system Public enterprises (PEs) Incomes policies External policies Macro policies & others Diversification of financial instruments (FIs); freeing interest rates (i); bankers’ scandal; partial reversal Attempt to reduce spending & increase taxes

PE accounts improve due to price increases

Military control of labor @)-market; agricultural support policies (ASP) curbed Real depreciation (RD); import quotas reduced, heavy X- subsidies (XS); some forex controls (FC) lifted Monetary restraint through CB credit controls, deregulation of internal markets

Environmental A series of environmental laws policies and regulations based on

“polluter pays” principle, with limited enforcement and open evasions

Interest rate policy under Interbank money and forex Central Bank (CB) control; markets with CB as participant; multiplication of FIs; gradual freed interest rates monitored domination of gov’t bonds (GBs) by CB

Lowering tax burden on High interest burden on budget; business, rentiers &high Treasury - CB conflict on incomes; value-added tax financing public deficit (PD) (VAT) adopted; higher

spending after 1984 Privatization (PN) agency set up; minority shares sold; debt financing of PE investments

PN seen as major revenue source; debt burden worsens PE accounts

Anti-L, anti-farm, prorentier policies; L militancy in 1988

Return to populism on wages & ASP

Crawling peg targeting RD; XS All FC lifted; fully liberal K- remain; imports liberalized, account; market-based exchange external borrowing by firms rate; real appreciation

&banks eased

Reduced control over PD, attempt at fiscal restraint in 1988

Indirect taxation on energy inputs: Energy costs accelerate over private producer prices while energy/output ratios am not affected

Control over PD & internal debt growth lost; CB loses ground on monetary policy

Ministry of Environment (ME) founded in 1991; efforts to administer an earmark environmental funding system under ME

(3)

Table 2. Major indicatorsfor subperiods, 1981-92 (index numbers (1977: 100) andpercentages)

1981-83 1984-87 1988 1989-92

GDP Growth, % pa. 5.0 6.9 2.1 3.7

Main macroeconomic aggregates as % of GNP* - CAD, 2.4 1.7 -1.7 0.3 - SC 14.6 15.7 20.1 16.3 - 1, 17.0 17.4 18.4 16.6 - PSBRc 2.9 2.2 1.7 5.3 - CAD, 0.5 0.8 -2.8 -0.7 - S, 19.0 18.1 20.7 18.9 - 1, 15.0 18.9 17.9 18.2 - PSBR, 3.1 2.2 1.2 5.3 Inflation, % p.a.t 31.4 38.8 68.3 60.5

Capacity utilization rate, %$ 88.5 95.1 97.6 96.9

Real wages 77.0 63.3 68.9 93.76

TOTAR 60.7 63.0 50.0 65.4

REERII 123.7 148.0 151.9 128.9

Real interest rates, %**

- With ref. to current inflation 8.W 12.9 25.9 9.53X

-With ref. to future inflation O.OVt 10.6 30.9 12.4$$

*Subscripts c and r in the following rows refer to current price and real (1987) price calculations. respectively. tin terms of WPI.

$Method of calculation is described in Boratav and Tiirel(l989). §For 1989-90.

uerms of trade for agriculture on the basis of implicit GDP deflators.

IlBased on Uygur (1991) for 1981-90 and on Central Bank end-of-year estimates for 1991-92, this is real effective exchange rate for a basket of foreign currencies. Its increase signifies real depreciation of TL and vice versa.

**Average bank lending rates based on Akkurt et al. (1991) deflated by WPI. ttRefers to 1982-83.

#Refers to 1989-90.

over the past decades. As will be discussed later, the stagnation and even erosion of overall investment rates are linked to substantial declines in public investment which is not made up by private investment. (See Table 5.)

- Current account deficit (CAD)/GNP ratios, when averaged over the subperiods, give the impression of an overall improvement in external balance. Despite the appearance of alleviation of the chronic deficit position after 1988 with three surplus years, annual figures exhibit erratic and easily reversible movements. Following liberalization, the capital account also became more volatile. - The corrective effort to stem the rising trend of PSBR in 1988 turned out to be abortive when inflation moved onto a new and higher (60% p.a.) plateau from one around 30% p.a. in absence of significant cost-push elements. Despite the “wage explosion” and increased public deficits, the rate of inflation did not accelerate after 1988, but settled on this higher plateau. Real appreciation of the Turkish lira (TL) apparently alleviated inflationary pressures.

-Wages and agricultural terms of trade indicators

for the last two subperiods reveal the extent of the post-1988 distributional changes - a drastic reversal of earlier trends which is studied in more detail in Boratav, Tiirkcan and Yeldan (1994). Although subject to sizable fluctuations, positive and sometimes very high interest rates prevailed throughout 1981 to 1992-93. Despite the appreciation in the last subperiod, TL remained depreciated in real terms, in comparison to the values in the late 1970s and in 1980.

Despite the comprehensive economic and institutional reforms, environmental protection has nevertheless been a neglected component of policy action in Turkey. Even though environmental concerns have been explicitly voiced in various policy documents, the country still lacks a sound environmental research and action program. Environmental policy-relevant research is still at its infancy; existing data are limited to air and water pollution in provincial areas, and is of dubious quality. On the other hand, the existing environmental policies are put into effect through a mix of pervasive but hardly enforceable controls and polluter charges which are neither comprehensive nor suffi- cient to prevent continuing environmental degrada-

(4)

Table 3. Decomposition of changes in external deficit, 1981-92 (as percentages ofpotential outpuf)*

1981-84 1984-87 1987-88 1988-92

Change in external Deficit’ -0.2 -1.3 -2.6 +2.3

External shock component +1.3 -2.8 -1.0 -1.7

(Terms of trade changes)$ +0.6 -2.8 +1.3 a.9

(World trade expansion) +l.O +0.8 -1.0 -0.5

(Interest rate shock) -0.1 -0.3 -0.1 -0.1

(Change in transfers) -0.2 -0.6 -1.2 -0.2

Debt burden component$ +o.s +0.4 +0.4 -0.7

Policy component -2.0 +1.2 -2.0 +4.7

(Increase in u)cfi +0.2 +0.3 -0.1 -0.0

(Import compression)” +2.6 +2.5 -1.3 +3.0

(Export penetration)** -4.8 -1.6 -0.6 +1.8

* figures may not sum up to totals because of rounding.

t Negative figures imply reduction in balance of payments deficit relative to potential output. $ Positive figures imply adverse changes.

8 Positive figures imply adverse changes. The negative value for 1988-92 is due to real appreciation of TL. I u: rate of capacity utilization.

I’ Negative figures imply import compression.

** Negative figures indicate export penetration due to reduced absorption. Source: Boratav, Tiirel and Yeldan (1994), p.7,

1981-92 -1.8 -4.3 -1.9 +0.3 a.5 -2.2 +0.5 +2.0 +o.s +6.8 -5.3

tion. In parallel with the general tenor of the post- 1980 reforms, Turkish environmental authorities (mostly with the support of the World Bank and other international donors) also tried to design market-based policy measures and thus shift from regulation to price mechanisms in the realm of environmental management. Most of these measures however have not been imple- mented yet, mainly due to the lack of political will and proper organizational structures to administer envi- ronmentally sound tax/subsidy schemes.

(b) Decomposition of changes in external andfiscal gaps

Constituent elements of changes in external and fiscal gaps in line with Gibson’s (1991) decomposition framework are presented below. Although there are quite diverse movements in the subperiods, only the picture for the entire period as shown in the last columns of Tables 3 and 4 is described here.l

Over the entire 12-year span, the reduction in foreign savings/potential output ratio (s,) by 1.8 percentage points was essentially realized by favorable external conditions: A decline of 3.8 percentage points in foreign savings is accounted by the “external shock” component, after allowing for the impact of the debt burden; 2.2 percentage points out of the 3.8 are due to the rise in service exports and other invisibles. Although in the model framework these latter categories are incorporated in the “external shock” com-

ponent, it should be noted that they are also responsive to policy variables. Nevertheless, the fact remains that adverse international developments which dominated 1973-81 were, as a whole, reversed in the following period.

On the policy front, the impact of import liberalization (leading to a 6.8 percentage points rise in s,) more than offset the positive impact of the export drive pro- ducing a 5.3 percentage point decline in sf. Hence, contrary to the widespread belief, economic policies, in toto, contributed to rises in external deficits by 2.0 percentage points. Had these policy choices produced an overall neutral impact on sr the Turkish economy in the early 1990s could have attained a modest, but per- sistent surplus in balance of payments instead of erratic movements.

Over 1981-92, a rise in 2.0 percentage points in PSBR is essentially explained by domestic interest rate shocks and changes in the domestic debt burden which account for a rise by 2.7 percentage points. The mildly favorable impact of external and policy components (-0.2 and -0.5 percentage points, respectively) account for the rest. Although the method of decomposition includes the impact of local interest rates in the “domestic shock” component, in the Turkish context it is essentially a policy-determined factor. One can conclude that the deterioration of public sector finances over time is, to a large degree, the outcome of costly domestic borrowing schemes, which toward the end of the period started to acquire the characteristics

(5)

Table 4. Decomposition of changes in PSBR, 1981-92 (as percentages ofpotential ouputJ*

1981-84 198&87 1987-88 1988-92 1981-92

Change in PSBRt -2.3 +1.7 -1.8 c4.4 +2.0

External shock component +0.3 +o. 1 +0.3 -0.9 -0.2

(Debt burden) +0.4 +0.3 +0.3 -0.8 +0.3

(Interest rate shock) -0.1 -0.2 -0.1 -0.1 -0.5

Domestic shock component (Debt burden) (Interest rate shock)

+0.3 +l.O +0.7 co.8 +2.7

-0.2 +0.8 -0.0 +o. 1 +0.7

+0.4 +O.l +0.7 +0.7 +2.0

Policy option component (Tax changes)$ (Distributional changes) (Increase in u) (OtherS -2.9 +0.7 -2.8 +4.5 -0.5 +2.7 -2.6 +l.O -2.7 -1.6 -0.1 Al.3 +O.l -0.5 -0.8

+O.O -0.0 +o.o 20.0 +o.o

-5.6 +3.6 -3.9 +7.7 +I.9

* Figures may not sum up to totals because of rounding

t PSBR stands for the investment-saving gap of the public sector relative to potential output. $ Increased tax effort produces a negative sign.

5 Impact of changing spending, tax and transfer patterns. The relatively large figure for 1988-92 is mainly due to rising current spending and erosion of PE surpluses.

Source: Boratav, Tiirel and Yeldan (1994), p. 9.

of Ponzi financing. The policy option components which accounted for many of the changes in successive subperiods seem to offset each other, leaving interest and debt shock as the major determining factors of increased public deficits.

Policy makers’ response to increased public deficits was an unwillingness to reduce transfer spending and undertake serious tax reform. There is a -2.4 percentage point tax effort component behind the modest -0.5 percentage point policy contribution to PSBR growth. But, as will be discussed below, this basically consists of (i) increased productivity of indirect taxation and (ii) the upward movement in wages and salaries during the last subperiod and the resulting increase in income tax revenue. No effort at reforming direct taxation took place. This led to a vicious circle of deficit finance via borrowing as the preferred alternative to meet expenditure gaps with all the adverse consequences.

The indirect taxation regime had important consequences for environmental policy making. It can be argued that in the absence of a comprehensive, direct regulatory system, price policy remains the only instrument to address environmental concerns. The rise in the cost of energy which increased by almost 50% faster than the industrial producer price in Turkey over the last decade is worth noting in this context. Thus, it can be argued that energy pricing policy suc- cessfully reflected the scarcity value of resources to the industrial producers. It could also be cited among

the so-called “win-win” policies of the World Bank in which policies for “efficient” income growth comple- ment those for environmental protection (World Bank, 1992). The technological input choice aspect of this policy has nevertheless yielded a rising trend in energy consumption both in terms of per unit of output and per labor employed in the manufacturing industry. despite the rising energy costs and a steep drop in real wages paid in the sector. On the other hand, polluter and/or user charges throughout the post-1980 period were neither sufficiently high nor strictly enforced to have a significant impact on the protection of environment.

(c) The state and the bourgeoisie: Corruption and rent-seeking under a “liberal” regime If we label as “rent-creating activities” all cases of economic decision making and policy implementation which create positions of advantage for individual businesses, as the economic regime moved further into liberalization, a broader the scope of rent-creation and rent-realization appears in post-1980 Turkey. This striking paradox created an embarrassing challenge to partisans of orthodoxy who had been advocating liberalization on the basis of its favorable distributional consequences by lowering rents of all sorts emanating from a large, protective, interventionist state.

The problem with the utopian neoliberal position is that under the structural and class characteristics of

(6)

less-developed capitalism, there are objective limits to the shrinking of the state and the very process of liberalization creates new rents. In this context it should be noted that in most Third World countries the bourgeoisie itself is a creation of the state and this histori- cal phenomenon has created cultural, sociological and economic traits which do not disappear with changes in the policy model. The very process of rent-seeking emanates from the bourgeoisie, not from the state per se and this is a well-known daily fact in the Turkish scene which only academic liberals ignore. The major areas of rent-seeking and rent-creation under the liberal model have been (i) illicit appropriation of export subsidies through fictitious exports, (ii) speculative urban and touristic rents, (iii) privatization and stock exchange operations, (iv) government tenders, (v) fiscal incentives and bailing out of firms and banks in distress. These are reviewed in some detail in Boratav, Ttirel and Yeldan (1994). Although it is difficult to quantify, indirect evidence suggests that the relevant magnitudes (i.e. the rents or “spoils”) involved during the “liberal” phase may have been substantially larger than those under the ‘*protectionist/interventionist” phase. The mere fact that there has not been any period in the history of republican Turkey when corruption and economic scandals have so heavily dominated the public scene as the present one is significant in this context. Far from being transitional phenomena, rent- creating categories either persist or, as old patterns die out, new ones emerge.

There was, moreover, another new element even in the process of rent-creation under the liberal regime: Decision making and implementation on rent-allocat- ing activities definitely shifted into the prerogative of the political layer of the state structure as the upper bureaucracy was pushed increasingly into a passive position. This was a conscious policy move and was carried out under the banner of “less government by eliminating the bureaucratic fetters in economic decision-making.” During the import-substituting phase, the basic function of the bureaucratic layer has been not the elimination of “rents,” but rather guaranteeing neutral and/or acceptable conditions among rivaling claims with respect to the distribution of rents. Although this state of things was never fully satisfactory for the political “class” and their constituencies within the business community, it was nonetheless instrumental in creating a significant degree of stabil- ity.

This modus vivendi which was gradually being undermined by uneasy coalition governments since the mid- 1970s was totally destroyed during the second half of the 1980s. A number of functions traditionally resting with the bureaucracy shifted to the political layer, e.g., directly to the Ministers or the Prime Min- ister. Key posts under the Ministry of Finance, SPO and the Central Bank, which had always been kept out of the reach of “outsiders” were now given to experts

“imported” from abroad with no experience in Turkish public administration and who acted merely as exten- sions of the political “class.” The traditional bureaucratic layer was gradually rendered ineffectual. More- over, the gradual erosion of the real salaries of the upper bureaucracy had a destructive impact on the quality of public administration. The eventual paraly- sis of the bureaucracy resulted in arbitrariness in rent creation and realization: Personal connections of busi- nessmen with ministers gradually became the determining factor in access to the “spoils” emanating from the government.

These new developments led to acute tensions and inner conflicts within the bourgeoisie as well as between certain business groups and the state particularly after the 1985-87 economic boom came to an end. Since all segments of the business community could not have equal access to the highest echelons of the political layer, it was natural that these orienta- tions, in the end, created more adversaries than sup- porters and led to the emergence of a strong sense of uncertainty and dissatisfaction within the bourgeoisie at large during the late 1980s.

2. WORKINGS OF THE ECONOMY (a) Accumulation patterns

Index numbers of public, private and total gross fixed capital formation (GFCF) are presented in Table 5, taking 1977 as the reference year. The following points can be emphasized:

- In constant (1987) prices, the overall 1977 GFCF level was reached in 1986 and by 1992 exceeded the former magnitude by 30%. Due to GDP growth, this results in declining IGDP ratios (calculated at 1987 prices) during 24.5% to 16.8% during 1977-92. For the public sector, 1992 figures are around the 1977 levels. Private GFCF reached the 1977 level also in 1986 (1989 if we exclude housing investments). Annual growth rates for total and private investments were slightly lower than 1.8% and 3.5% p.a. for the 15-yearper- lod beginning in 1977.

-There were two upward spurts in private investments: The first was due to the boom in housing investments which attained a 36% p.a. growth rate 1985-88 and the second in was a one-year “explosion” in real manufacturing GFCF by 64% in 1990. The level attained in 1990 was preserved during the following two years. It is very difficult to explain the second spurt with reference to conventional arguments (e.g., interest and inflation rates, crowding-in or crowding-out, acceleration factors, etc.) and this raises doubts about data reliability. - In terms of sectoral distribution, despite the

1990 boom in private manufacturing investments, dramatic declines in public GFCF pushed real total

(7)

Table 5. Index numbers (1977=100) of GFCF by sectors (1977: loo) 1981 1984 1989 1992 GFCF, 95.2 87.5 89.4 101.4 A+M+M 105.2 69.9 38.6 40.3 Infrastructure 92.0 103.9 126.4 122.6 Services 75.6 91.6 130.3 215.5 GFCF, 62.7 83.0 143.5 166.9 A+M+M 55. I 76.8 74.9 117.6 Infrastructure 75.0 104.4 114.2 188.5 Services 68.7 83.0 257.0 230.3 GFCF, 78.9 85.5 113.4 130.4 A+M+M 80.9 73.3 56.4 77.8 Infrastructure 88.3 103.8 123.6 123.6 Services 71.1 86.2 212.2 225.3

* Subscripts g, p and t refer to “public”, “private” and “total.” Rows below GFCF, and GFCF,, are public and private sec- tom1 GFCF figures. A+M+M: agriculture + mining + manu- facturing; Infrastructure: energy + transport and communica- tions: Services: all other sectors.

manufacturing GFCF 20% below its 1977 level. Indeed, in 1992 total real GFCF was below the 1977 level in agriculture, mining, manufacturing and energy sectors. It is, in our view, this dramatic shift against the “productive” sectors that lies behind the decline in potential growth rate. - Crowding-out theses of private by public investments are not validated in the Turkish case by econometric analysis. Financial crowding-out is irrelevant in this controversy because it refers not to public investment per se, but rather to public deficits covered by borrowing whatever the underlying cause. Our own data suggest tradeoffs between private investments in housing versus manufacturing, rather than the public/private dichotomy.4

Declining rates in (and drastic changes in the sectoral distribution of) capital formation was not accompanied by higher efficiency of investments. The changing path of capital formation with no discernible improvement in output/capital ratios5 merely resulted in lower potential growth and a shift toward relative stagnation. Economists who share the foregoing diag- nosis are in agreement that a more active state is required for moving into a dynamic path of develop- ment.

(b) The pattern ofprice movements (i) The wage cycle

A grassroots workers’ movement starting in 1988 combined with the catastrophic electoral setback of 1989 for the ruling MP changed the industrial-social scene altogether. Elements of conventional populism have reemerged within, however, an essentially different overall environment: Turkey, unlike the 197Os, is now a major exporter of industrial products and, as will be discussed below is critically vulnerable to wage movements. The economy is facing a serious fiscal constraint which is also highly sensitive to upward changes in public sector wages and salaries.

Table 6 provides a picture of the wage cycle of 15 years. By 1990 the preceding peak (1977 or 1978) for real wages was attained, but the W/Y ratio was substantially behind the all-time peak due to productivity improvements which took place during the period. The magnitude of the wage explosion of 1989 and

1990 can be observed by the 1990/1988 ratio for real wages: 72% or 51% increases in real wages in the PE sector and in manufacturing industry respectively in two years can only be characterized as an internal shock to the economy. Turkish experience is striking both in terms of the length and magnitude of the wage erosion and of the speed and magnitude of the “correc- tion.” As the relationship between competitiveness and wages further on shows, the Turkish case also provides a lesson on the futility of relying on wage compression alone as a medium-term policy variable for achieving competitiveness. Public sector wage increases were significantly higher than those of the private sector which tried to defend its profits by layoffs, part-time schemes, subcontracting, fictitious dissolution and reestablishment of companies. 19900988 real wage ratios for public and private manufacturing industry were 1.76 and I .42, respectively. The outcome was further deterioration offinan- cial indicators for PEs after 1988 which facilitated the arguments of the pro-privatization lobby.

Turkish data on manufacturing industry also enable us to look into the determinants of the wage share. Apart from the simple relation between productivity and real wages, another way of looking into the problem is through an investigation of the determinants of mark-up rates: Under assumptions of mark-up pricing, wages I value-added ratios are, by definition. determined by the mark-up rate, and the ratio in v&e terms between nonlabor input costs (C’) and wages (W). Within such a framework, the factors which will affect W/Y downward are: increased mark-up rates. increased capital intensity of production under unchanging relative prices (constant real wages in terms of nonlabor input costs), and a decline in real wages (in terms of input costs), that is relative increases in intermediate costs (in terms of labor costs).

(8)

Table 6. Index numbers of real wages and wages/value-added

(W/u) ratios, selected years*

Years W, W,, WJY, WdYv r(%o) CIW

1976 100.0 100.0 34.3 29.8 .358 5.601 1977 113.2 108.3 37.1 35.1 .305 5.065 1980 96.9 81.9 30.7 27.6 .359 6.318 1983 85.1 80.1 24.8 26.0 .316 8.015 1986 70.0 68.0 16.1 18.2 .381 10.832 1987 74.9 76.6 17.3 17.4 ,382 11.404 1988 74.0 74.6 15.4 16.7 .398 11.577 1989 103.7 90.1 19.0 19.7 ,363 10.246 1990 127.1 112.8 21.7 21.8 .408 7.815 TIP .556 .623 ,415 .476 1988/P ,587 ,689 .415 ,476 19900988 1.719 1.517 1.409 1.305 1990/P 1.009 1.042 ,585 .612

W,,: Real wages in PEs, Wmr: Real wages in total manufacturing. T: All-time trough in real wages, P: Pre-1980s peak. r: Mark-up rates in private manufacturing. C/W: Nonlabor input costs relative to wage bill in private manufacturing. The deflator used in real wage calculations is based on wholesale price series of State Institute of Statistics (SIS) for 1981 onward and of the Treasury for preceding years. WJY,,, and W,,JYmp refer to total and private manufacturing sector in current prices, respectively.

The manufacturing industry data for 197690 suggest that the period covered can be divided into years when, first, stable mark-up rates + changing C/W

ratios occurred, and second, when relatively substantial changes in (r) together with minor or significant changes in C/W ratios are observed. 1977, 1981, and 1983-85 are years when r dropped significantly to a lower level to recover immediately afterward, sometimes to a higher plateau. The other years represent, roughly speaking, the first case above.

The data further suggest that productivity and real wage changes are the main determinants of variations in C/W ratios. Relative nonlabor input prices, chang- ing input coefficients and industrial coverage provide additional, but less important explanatory factors.

During a period of radical change in relative prices such as 1976-90, mark-up rates have been strikingly stable. On the basis of available Turkish data, one can propose three possible determinants in explaining changing mark-up rates: rates of capacity utilization, rates of protection for the relevant industries and rates of change in real wages.

The analysis in Boratav, Tiirel and Yeldan (1994) shows that until 1976, real wage growth rates seem to be more closely (and inversely) related with mark-up rates and capacity utilization rates seem to be another explanatory factor. The latter variable seems to affect

(r) during 1974-77, the last years of the gowth cycle of

the 1970s. During 1978-82, real wage declines seem

to explain the rise in (r) since this is a period of very low rates of capacity utilization. 1983-85 seems to have been an exceptional phase since it incorporates a decline in (r) together with lower real wages and higher capacity use. This may be the only subperiod where import liberalization under a stagnant conjunc- ture may have affected the erosion of profit margins. Higher capacity use in 1986-88 and 1990 and real wage movements in opposite directions in 1988 and 1989 seem to be the determining factors behind mark- up movements (continually rising rates from 1984 except in 1989). The absence of any strong link between rates of protection and mark-up rates is striking except, as we have noted, for 1983-85. There seems to be no solid empirical ground for linking the so-called rents of protection and the trade regime. We return to this question when we investigate sectoral relative price movements below.

It should be clear that oligopolistic mark-up pricing has direct consequences for environment via the choice of input mix. For example, there is evidence that cost savings achieved from repressed wages and lagging material input prices (of the public enterprises) served as a buffer against rising energy prices in the private manufacturing sector (Yeldan and Bakan, 1993). Consequently, in an oligopolistic environment private producers could exercise pricing strategies based on cost-plus-mark-ups and when prices of other factors of production lagged behind the price of energy, oligopolistic producer surpluses could be maintained, and the rise in energy prices did not necessarily result in substitution away from energy inputs. The pressure of this setting on the environment is potentially enormous, and is further compounded by the fact that the present technology in electricity production relies heavily on the use of low-grade coal.6 We reflect more on this issue inspection 4(c) below. (ii) Agriculture’s terms of trade

Table 7 produces the findings on Turkish and international terms of trade for agriculture (TOTA, and TOTA,, respectively). It shows the dramatic regres- sion in 197688 and the substantial recovery in 1989-90 in parallel with the “wage explosion” and as part of the overall process of the reemergence of populism. Table 8 provides hints on the role of the policy tools (i.e. the relative magnitude of support purchases and relative support prices) as a determinant of TOTA. Differential (i.e. flexible vs. mark-up) pricing behavior of agriculture and industry under specified macro policies and in response to productivity may provide an additional explanation of TOTA movements. On possible other determinants, differential income elas- ticities of demand and different income growth for the two sectors cannot provide an explanation of agricultural price movements of the 1980s in view of the realized growth of incomes and acceptable elasticity values.

(9)

Table 7. Inlernal and inremational terms of trade for agri- culture: index numbers, ratios and rate of change (%J

Years TOTA, TOTA, T/W G,(s) G,,

1973 111.7 174.0 .642 - - 1974 119.8 162.2 .739 +7.3 -6.8 1975 116.9 143.0 ,817 -2.4 -11.8 1976 100.0 100.0 1.000 -14.5 -30.1 1977 99.5 89.5 1.112 a.5 -10.5 1978 96.7 104.1 ,929 -2.8 +16.3 1979 81.4 104.5 ,779 -15.8 +0.4 1980 67.3 137.2 ,491 -17.3 +31.3 1981 65.4 118.3 ,553 -2.8 -13.8 1982 56.7 90.3 .628 -13.3 -23.7 1983 59.2 99.3 ,596 +4.4 +lO.O 1984 66.1 92.3 ,716 +11.7 -7.0 1985 63.7 79.5 ,801 -3.6 -13.9 1986 62.1 71.4 ,870 -2.5 -10.2 1987 59.0 73.3 ,805 -5.0 +2.7 1988 49.8 80.8 ,623 -15.6 +9.1 1989 ’ 57.8 81.9 ,706 +16.1 i2.4 1990 70.7 73.1 ,967 +22.3 -10.7 1991 66.2 69.5 ,953 -6.4 -4.9

Table 8. Relative magnitude of support purchases (9;) and relative support prices (ratios between index

numbers) Years SP/A VA p/p 1976 14.7 1.000 1.000 1977-79 12.2 ,752 ,778 1980-86 8.6 ,673 ,867 1987 7.3 ,723 .856 1988 5.5 ,797 1.029 1989 5.7 ,826 ,944 1990 7.6 1.240 1.785 1991 12.0 1.143 I.268 1992 13.2 1.166 1.167

Notes: Calculations based on SPO and SIS data. SP: Value of support purchases by public agencies. AVA: Agricultural value-added. P, : Index number of weighed support prices for wheat, tobacco, sugar beet, cotton, sunflower seed, hazelnuts and tea (only the first three are considered for 199&91. and the first six for 1992). P: Wholesale price index. P, : Index number of implicit deflator for agriculture as calculated from GDP series.

Notes: TOTA, for 1973-89 is from Ttirel(l99 1) and updated for 199&91 based on the ratio between UNCTAD’s price indices for agricultural and manufactured commodities in world trade. TOTA, is based on GDP implicit deflators. G, and G, are annual growth rates for TOTA, and TOTA,.

As for the role of international terms of trade movements as a possible determinant of TOTA for Turkey, the question is important not only in its own right, but also for understanding the degree of freedom of domestic support policies vis-ci-vis world market para- meters. The overall impression one can get from Table 7 is that there is not a high degree of dependence of TOTA, on TOTA,: Out of the 18 years covered in the table, the direction of change is the same (both + or- signs) in only 10 years. It seems however, that following the price shock of 1980 - a year where, paradox- ically, there was a significant improvement in TOTA,-, internal price movements seem to fall more in line with international terms of trade: Divergence in direction of change in TOTA, and TOTA, occurs in four out of seven years during 1973-80, but only four out of 11 in the post-1980 years. Fluctuations around the mean during the first phase were relatively speaking, significantly higher than those during the latter period.

One can conclude that international price movements had some, and increasing, impact on local terms of trade; but it was not a dominant factor. By the end of

the 198Os, the ratio between local and international terms of trade was still significantly below the parity of 1976, and although 1987-89 and 1973-75 averages were nearer each other, Turkish farmers were still losers compared with world exporters of agricultural commodities. It also turns out that if local terms of trade had followed international prices during

1981-86, the relative price deterioration of Turkish farmers would have been significantly more serious.

The adjustment pattern of farmers in response to relative price changes is another crucial theme for debates on support policies. It is clear that the deterioration of relative prices affected farmers’ investments adversely during the 1980s. By the end of the decade. private investments in agriculture in real terms were less than one-third of the level attained in 1976. Public investments made up some of the deficiency, but still the overall picture is, at best, one of stagnant capital accumulation in agriculture. This would be expected to have adverse consequences for labor productivity. It did not happen during 1976-88, the trough year in terms of TOTA, average labor productivity growth is 1.3% p.a. - a rate which is, however, lower than that observed during the preceding 15 years.

To make up for the adverse impact of investment stagnation on output, farmers increased the intensity of land use in certain areas, but extended cultivation into marginal areas and/or undertook deforestation practices. Thus, as a percentage of the total cropland,

(10)

Table 9. Sectoral tenm of trade movements and the trade regime: index numbers, 1976=100 and shares/rates in percentage Years 1979 1980 1983 1988 1989 1990 1991 Ind. Agr. Serv. Const. Trade T&C Finance Housing Lib. Pr. Gov’t E, (X+M)IY (X+M)JY NP 111.1 90.5 104.6 130.8 93.7 102.4 88.0 195.4 100.6 101.0 83.8 9.3 11.1 - 116.7 123.6 138.5 78.5 73.1 68.9 110.3 109.2 104.9 116.0 119.8 153.0 112.9 115.1 108.0 108.2 121.7 114.9 76.9 96.0 151.0 125.0 124.5 85.6 95.7 98.6 82.8 82.8 61.5 58.5 111.7 153.6 169.7 15.5 24.3 27.8 13.9 15.8 16.4 129.0 119.0 59.0 131.2 75.9 104.3 134.6 109.5 111.5 133.3 69.4 80.5 81.8 154.6 24.9 16.1 - 118.1 117.9 83.4 80.0 105.5 107.6 136.7 147.4 101.9 100.4 109.8 110.8 160.9 202.1 64.8 72.9 80.0 79.7 106.8 117.0 134.4 143.5 23.3 23.2 17.3 16.2 28.7 28.3

* The first 10 rows are index numbers of sectoral relative prices from GDP series. (Each sectoral price index is related to the GDP’s implicit deflator.) Rows 4-10 are subsectors of the service sector. The last four rows are effective real exchange rate (TL per dollar, index number), trade/GNP shares in nominal terms, in terms of REER and weighted average of nominal tariff rates (in percentages) respectively. For Er see Table 2. NP, taken from Krueger and Aktan (1992) is an unweighted average rate of protection except for the 1990-91 figures which are weighted averages taken from Togan (n.d.). Research assistance by N. Solakoglu is acknowledged.

the share of irrigated land increased only marginally from 7% to 9% over 1980-89. With shortened periods of fallow, the nutrient capacity of arable soil declined and together with forest clearing, exacerbated the extent of soil erosion.’

Complicating these problems further was the fact that more intensive agricultural practices resulted in the declining immunity of many crops to pest attacks. A typical response to such threats was to increase pesticide dosages or the frequency of spraying. This would offer temporary relief but in the meantime would only hasten the trend of pesticide resistance. In the process, farmers have become trapped on what the RFF (1994) refers as “the pesticide treadmill,” cycle of ever-growing dependence on chemicals that offers diminishing returns with consequent increased water pollution, while necessitating higher cash expendi- tures.

(iii) Sectoral relative prices and surplus redistribution

protection (NP) and external trade/GNP ratios (X+&f/y] are indicators presented in Table 9 represent- ing various aspects of the trade regime or the degree of “opening-up” of the national economy. On the basis of the foregoing changes on the trade regime and “opening-up,” incorporating significant import penetration as construed from Table 3, an orthodox neoclassical approach would expect relative prices between nontradeable / tradeable sectors (roughly represented in our table by the relative sectoral price for “services”) and relative prices for industry and trade (incorporating “rents of protection” and commercial margins, respectively) to decline significantly between the early and late 1980s. An improvement for relative prices in labor-intensive agriculture would then take place. 1988 can be set as a reference point by which all these predictions can be tested in view of the real depreciation by 5 1 %, a threefold increase in the share of the external trade from 1979 and a decline by 54% in nominal protection.

Table 9 provides findings on sectoral terms of trade Table 9 shows that orthodox predictions did not movements. The critical issue to be discussed is the take place. The predominantly nontradeable service following: Is the trade regime a crucial determinant of sector’s terms of trade improved by 5% during sectoral prices and, hence, of income distribution? Or, 1976-79 and remained stable thereafter. Industrial and alternatively, do internal and autonomous changes in agricultural relative prices moved in opposite direc- income distribution affect sectoral relative prices with tions, with an improvement in the former and a deteri- the trade regime playing a secondary role, if any? oration in the latter. Trade margins register a signifi-

(11)

Table 10. Indicators on the redistribution of the surplus* Years 1976 1977 1979 1980 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 RIP,, 52.6 54.1 60.4 67.6 50.1 56.3 49.9 48.2 69.5 RIP, 13.4 22.1 53.7 42.0 46.0 39.0 45.2 41.5 42.2 47.3 63.3 - Fin/GDP TrIGDP 2.2 15.3 1.9 15.4 2.1 15.3 2.0 16.5 2.3 18.3 2.2 19.2 2.2 19.0 2.4 17.9 3.0 20.3 3.3 20.3 2.8 19.7 3.2 19.5 4.2 18.9 4.3 19.1 R/GDP R(r) 2.0 100 6.8 393 7.9 488 8.5 579 8.8 677 8.2 695 9.8 857 10.3 929 8.7 859 P(br) IO0 247 329 637 618 969 1310 I007 1525

* For symbols and concepts, see text. The first three columns are percentages, the last two columns are index numbers. RIP,,,, and RIP, refer to 500 largest firms in Turkey and to industry in Eskisehir region. respectively. The table is an updated and adapted (to new GDP series) version of Boratav (1991, Table 1). R/GDP and R(r) are estimated anew on the basis of Temel’s (1991) findings to which public sector interest payments are added.

We have already analyzed industry / agriculture price relations and therefore only discuss the factors which may have led relative prices for the service sector to violate neoclassical expectations. A very simple answer could be that the components (subsectors) of the heterogenous service sector incorporate different internal distributional dynamics, each of which have affected pricing behavior in a discordant fashion. The composite relative price for the service sector is, therefore, the outcome of what has happened within each of its components. Internal distributional dynamics emerge as the basic determinant in this respect, whereas the trade regime may act essentially as an indirect one.*

Sectoral terms of trade movements illuminate an important aspect of secondary relations of distribu- tion: Relative price movements of trade and finance vs. industry may provide some evidence with respect to the internal division of the surplus. Here, we investigate this aspect of the redistribution of the surplus between financial, commercial and industrial capital and rentiers more directly. The indicators used are the share of interest costs in industrial gross profits (RIP), the share of finance and trade in GDP (Fin./GDP & Tr./GDP respectively), the share of interest revenues in GDP (R/GDP), interest income in constant prices [R(r)], and (v) profits of the banking sector in constant prices [P(br)]. The relevant findings are presented in Table 10.

The first two columns of the table show that with respect to the internal division of the surplus, a dramatic progression of interest burden took place at the

cost of industrial profits during the 1980s. Financial capital and rentiers shared the benefits of “financial liberalization.” It should be noted that R/GDP and R(r) figures underestimate the size of interest revenue since they cover only those from time deposits, foreign exchange deposits, government bonds and Treasury bills but exclude interest revenues from investment funds managed by other financial institutions and from the new financial instruments of recent years.

Following the arguments in Akytiz ( 1991) and on the basis of casual observations, we can hypothesize that the typical nouveau n’che whose revenues are mainly of speculative character and based essentially on the ownership of financial assets and/or real estate have a low propensity to invest, exhibit extreme versions of conspicuous consumption and have both the capacity and inclination to switch resources abroad should appropriate conditions emerge. Their increased domination of the economic scene has serious, unpredictable and most possibly adverse consequences. The perception of their separate identity is important because it will be primarily this group that is targeted for any serious fiscal reform during the coming years.

(c) Destabilizing processes: thefiscal andpnancial systems

(i) Fiscal strains

The findings in Section 1 may suggest that Turkish fiscal deficits were not as serious as that experienced by a number of developing countries during most of

(12)

the 1980s. On the other hand, after the financial col- lapse of 1982 financial liberalization continued at a relatively smooth pace and by the early 1990s had attained an advanced level both internally and exter- nally. Despite these satisfactory appearances, the interrelationship between the fiscal and financial systems had started to create a number of tensions which, by the early 1990s. were leading to serious destabilizing processes.

The following points concerning the changing structure of the fiscal system and the factors leading to fiscal strain are worth emphasizing:

- On the revenue side, it is not the changes in the overall tax burden, but rather those in its structure which have been the characteristic feature of the period covered. Although agricultural incomes were not satisfactorily taxed and payroll taxes continued to be the most productive component, Turkey until 1984 was gradually moving toward a unified and progressive pattern of direct taxation with a direct to total tax ratio of around 57% in

1981. Tax reform in the 1980s reversed these developments. The income tax became drastically less progressive and more fragmented; dividends, interest income on bonds and government bills, realized capital gains on real estate and financial assets were excepted from taxation. Rental and intellectual property incomes and interest revenue on bank deposits were taxed at low and flat rates. Although the corporate tax rate was raised, effective rates were reduced by a variety of exceptions and incentive schemes. On the other hand, the introduction of value-added tax (VAT) increased productivity of indirect taxation, contributing to a

17 percentage point decline in the direct to total tax ratio during 198 l-92.

- On the expenditure side, interest payments on domestic debt soared by about four percentage points of GDP during 1979-92, due to ‘financial deepening” under high inflation. Public current spending was in a trough in 1988 mainly due to erosion of salaries, but upward adjustment of salaries and increased resource flow into military / security services pushed it upward again, bringing it to 12.2% of GDP, approximately two percentage points higher than that of the 1979.

- Stagnation in tax revenues and increased transfer and current spending brought pressures for a downward adjustment of public investments. In relative terms, public investments were nearly halved during I98 l-92 with a heavy incidence on productive sectors. Adjustment in Z, was not suffi- cient however for improving overall balance. Both Zp-S, and PSBR to GDP ratios rose considerably after 1988 reaching 7.2% and 10.4% respectively, in 1991.’ On corrective actions, the government opted for palliatives during 1988-92; e.g., increased rates of monetization, further downward

pressures on Ip and stepping up the privatization campaign.

(ii) Towardfinancial misadventures?

There is no doubt that in terms of conventional cri- teria “financial deepening” has taken place (Cinder and Tiirel, 1993, Table 25). The question is not whether such a process has actually taken place, but rather whether it has resulted in increased efficiency of the financial system.

By the early 1990s the multiplication and diversification of new financial assets and instruments had taken a chaotic form: two-thirds of all new issues were being realized by the banking sector, the rest within the capital market. To overcome the increased cost of holding deposits, banks started to devise a number of new instruments exempt from compulsory liquidity requirements on deposits. What resulted was the emergence of a number of financial instruments offering (under inflation rates around 60%) positive real interest rates for almost perfectly liquid holdings (Yeldan, forthcoming; Sak and Yeldan, 1993). These may be interpreted as flexible methods for overcoming the constraints imposed on banks by the Central Govem- ment. An important by-product was, however, increased resource flow into idle financial assets by firms. A breakdown of the aggregated balance sheet of the 500 largest industrial firms shows that the share of nonindustrial revenues in total revenues averaged 45.0% in 1991-92 in contrast to 17.5% in 1982-83.1° Most of the former consisted of short-term returns on risk-free financial instruments and represent the unnat- ural situation of firms increasingly acting as rentiers.

As will be elaborated below in the discussion on the role of banks with respect to hot money movements, even banks have started to act increasingly as institutional rentiers. The share of gross bank revenues or profits from capital market and forex operations and from government bond holdings can be used as an indicator of the phenomenon. During the first half of 1993, gross revenues and profits from these three sources constituted 47.6% and 42.3% of the total.” The share of credit in the total assets of the banking system declined from 51.3% in 1980 to 36.5% in 1992.

On discussing the efficiency impact of financial liberalization, a distinction between allocational and technical efficiency will be useful. A narrowing of rate of return differentials between firms and industries would indicate an improvement in allocational efficiency. Another indication of improvement would be a reduction in the rates of nonperforming credits, loan defaults and bankruptcies of financial institutions.

During the 1980s the reallocation of credits of the banking system away from agriculture and industry toward construction, housing, transport, tourism and domestic trade does not signify a move in response to differentials in social rates of returns between sectors.

(13)

Within industrial subsectors, data for the private sector in the 500 largest industrial firms that coefficients of variation of profitability did not decline during 1982-91 (ijnder and Ttlrel, 1993). With respect to the latter indicator, although data coming from banks are unreliable, there is no doubt that financial liberalization was accompanied by higher rates of loan default Cases of bankruptcies or government-enforced merg- ers to prevent failures were common during the whole period.

Two indicators for measuring the technical efficiency of the financial system are proposed: costs of intermediation and real costs to the borrower. Figures on net spread in Akkurt et al. (1991) show that there had not been a permanent improvement in either. 1982 was an exceptionally chaotic year, yet it registered the largest net financial margins. During the relatively controlled period of 1983-87, the average net spread is 8.3 percentage points. The three years following the deregulation of interest rates in 1988, witnessed a significant widening of the spread. A more detailed breakdown of the components of the net spread shows that it is not labor costs, but rather the upward movement in management costs and net profits of the banking system which have contributed to the phenomenon. This is, in our view, persuasive evidence that Turkish financial liberalization did not correspond to a move toward a more competitive environment, but rather vice versa.

Finally, let us note some implications of external financial liberalization. Rentiers, firms and banks are the agents behind hot money movements which have dominated the capital account of balance of payments and external debt growth from the late 1980s onward. Rentiers shift between FX and TL assets and engage in currency switching, capital flight and its return. Firms shift between borrowing in TL or FX locally and/or abroad. Banks - under current conditions - borrow abroad and lend domestically. The relevant rates of return affecting these agents is discussed in Boratav (1993). Except for 1989, the direction of short-term capital movements in response to interest rates and nominal exchange rate changes conform to the expected behavioral pattern.

The magnitude of unrecorded capital movements can be estimated by calculating the divergence between external debt growth and the cumulative current account deficits after correcting for reserve movements, errors and omissions, changes in cross exchange rates and misinvoicing of exports and imports. A study” on hot money movements, estimates that (after all aforementioned corrections) during 1989-91, $3.3 billions of unrecorded capital out- flow had taken place. There are difficulties with the underlying economic mechanisms, but there is no doubt that a liberalized regime of capital movements and erratic interest/exchange rate movements have been building an important element of instability into

the external balances of the economy.

This is an extremely unpredictable environment. The 1989-92 experience shows the serious problems confronting a developing economy which decides to move into full external and internal deregulation in the financial system under conditions of high inflation: “The spectre of capital flight” becomes the dominant motive in policy making and creates commitments to high interest rates and overvaluation. The links of these two policy variables with the real economy, i.e. investment rates and the current account of the balance of payments, are severed. High credit costs not only discourage fixed investment, but also lead to a pattern of factor use which runs counter to environmental protection objectives. When adverse impacts particularly on the current balance become too destabilizing, real depreciation comes on the agenda which, however, needs to be matched by further upward adjustment in the rate of interest if currency switching or capital flight is to be restrained. Instability in exchange and interest rates creates feedbacks which lead the economy further into instability.

The rational solution is to reimpose controls on capital movements. There are, however, strong elements of irreversibility when a liberalized financial system is concerned. Some degree of control on short-term borrowing abroad of banks, firms and public agencies can be reestablished, but rentiers are less likely to be con- strained and it may very well be their collective behavior which carries the gravest risks of financial misadventures.

(d) Trade issues

An exercise at decomposing the factors behind the impressive Turkish export performance concludes that regional composition and product mix effects were either negative or negligible throughout except for 1981-82 when the regional factor prevailed mainly by successful penetration into Middle East markets.” The residual component, as high as 80%, turns out to be the major factor behind the Turkish export performance.

It is this residue which reflects changes in the competitiveness of Turkish export sectors. Considering the fact that export growth of the 1980s was based predominantly on industrial goods which now constitute roughly 80% of commodity exports, one should look into changes in the degree of competitiveness of the manufacturing sector in analyzing the export performance of the Turkish economy. It is comprised, essentially, of movements in (i) the real exchange rate (including subsidy components) for industrial exports, (ii) real wages and (iii) labor productivity.

Table 11 shows the improvement in industrial competitiveness (IC) during 1980-88 which was, essentially, due to real devaluation and wage repression. Tbe productivity improvement starting from the

(14)

Table 11. Index numbers of competitiveness of industrial exports and its determinants (1976=100)* REER (1) Real Wages (2) Hourly Wages ($) (3) Industrial Competitiveness Productivity Indicator (4) [(4)/(3)*1001 1977 98.2 108.3 119.9 97.5 81.3 1980 111.7 81.9 111.1 69.4 63.0 1983 153.6 80.1 85.4 87.5 102.5 1986 168.0 68.0 70.3 115.9 164.9 1987 162.8 76.6 77.0 120.0 155.8 1988 169.7 74.6 74.9 130.1 173.7 1989 154.6 90.1 100.5 123.8 123.2 1990 134.4 112.8 161.6 141.6 87.6

* For REER see Table 2, Column 4 is labor productivity. The other columns are from Boratav, Ttirel and Yeldan (1994).

trough of 1980 takes place in an apparently respectable annual rate of 8.2%. It should be noted, however, that the peak to peak rate of growth in industrial productivity during 1977-90 was merely 2.5% p.a. and this performance was a relative to other newly industrializing countries (NICs), a dismal one? Com- paring late 1970s with late 198Os, annual growth of labor productivity in Spanish and Korean manufacturing industries was 5.5% and 7.4% respectively. Despite this inferior productivity performance, Turk- ish IC improved dramatically in comparison with the two countries. This improvement was based totally on wage and exchange rate movements. It is ironical, however, that all the positive factors behind the improvement in Turkish IC were reversed in 1989 and 1990. In a sense, the painful operation of adjustment realized during the past decade toward export orienta- tion was eliminated within two years.

It should also be noted that the Turkish export expansion might also have had harmful environmental consequences, since almost three-fourths of Turkish manufactured exports originate from sectors such as food, textiles and apparel and metallurgy, characterized either by sizable inputs from a hard-pressed agriculture sector or by high levels of pollutant emissions.

On policy options, productivity improvements are crucially dependent on the level and composition of GFCF and, despite some positive signs in recent years, this has been one of the weak spots of the 1980s. Sys- tematic wage repression is no longer on the agenda and the best one can expect on that front is a social pact which aligns wages to productivity movements.

On exchange rate policies, industrial exporters which are committed to opening up their financial systems are doomed to real appreciation of their currencies which effectively and quickly erodes their competitive power. When a certain level as an industrial exporter is attained, the proindustrial lobby is dis-

cemed by a commitment at least to a PPP-based exchange rate policy, if not to real depreciations so as to preserve or improve the competitiveness of the industrial sector. But exchange and interest rate policies with a view to inducing short-term capital inflows jointly create adverse impacts on the real side of the

economy: declines in IC and GFCF are inevitable. Lower GFCF undermines IC further by its adverse impact on productivity movements. Under these cir- cumstances, countries such as India, that can still pur- sue exchange rate policies based on current account targets are better placed than countries such as Turkey that have opted for the luxury of liberalizing their capital movements.

3. THE ENVIRONMENT AND THE ECONOMY (a) Statement of the problem

A fully-fledged macro model of environment-economy interrelationships for Turkey necessitates the elaboration and the quantification of the following links: depletion rates of resources and its relation to output, “production” of waste and its impact on environmental quality via pollution stocks, resource build- up through capital investment, removal of pollution via clean-up expenditure, and welfare effects of resource and pollution stocks (see Pezzey, 1992, p. 7). Although technical coefficients in the first four links are estimable in principle, it is hard to establish quantitative relations between the loss of environmental quality (itself a function of pollution stocks together with natural resource stocks) and the ensuing output loss on the one hand, and conversion of waste flows into pollution stocks on the other. The final link is a virtually unexplored area in terms of applied research in Turkey.